NBA Positions By The Numbers – Usage, Playmaking and Possession

Mar 1, 2015; Houston, TX, USA; Houston Rockets guard James Harden (13) and Cleveland Cavaliers forward LeBron James (23) during the game at Toyota Center. Mandatory Credit: Troy Taormina-USA TODAY Sports
Mar 1, 2015; Houston, TX, USA; Houston Rockets guard James Harden (13) and Cleveland Cavaliers forward LeBron James (23) during the game at Toyota Center. Mandatory Credit: Troy Taormina-USA TODAY Sports /
facebooktwitterreddit
Mar 1, 2015; Houston, TX, USA; Houston Rockets guard James Harden (13) and Cleveland Cavaliers forward LeBron James (23) during the game at Toyota Center. Mandatory Credit: Troy Taormina-USA TODAY Sports
Mar 1, 2015; Houston, TX, USA; Houston Rockets guard James Harden (13) and Cleveland Cavaliers forward LeBron James (23) during the game at Toyota Center. Mandatory Credit: Troy Taormina-USA TODAY Sports /

Last week, I started this intermittent series of Positions By the Numbers with a look at shot locations, to show that its own complex way, the NBA still hews pretty close to basketball orthodoxy. Positional revolution or now, players tend to shoot from areas predicted by nothing more complex than a warm-up five-on-oh secondary break drill.  Today, I’m going to move on to quickly examine generalized roles of players in NBA offensive schemes. As with last week, the categorization of players into positions is fraught with imperfection and overlap[1. Did Goran Dragic or Eric Bledsoe play point guard when they shared the floor in Phoenix? In that case, I split the baby and called them both point guards, but that only illustrates the difficulty inherent in mapping the slightly artificial construct of positions onto outlive playing styles.], but since we’re dealing with the NBA as a whole, we can live with a little bit of a fudge factor.

These positional averages are a product and extension of the True Usage methodology for assessing overall offensive roles. This accounting allows for comparisons of scoring and playmaking usage as well as allows for a method for better capturing turnover propensity than traditional turnover rates.[1. TO% is usually calculated as TO / (TO + scoring attempts). Naturally, this punishes playmakers relative to scorers as the only way to “lower” your turnover percentage is to simply shoot more. It’s simply illogical to use a method for capturing looseness with the ball in which the players tasked with handling the ball the majority of the time – point guards – show up as among the worst.]

  • Scoring

Similar to traditional usage rate, Scoring Usage is an estimate[3. For ease in calculation I’m using approximations for both scoring attempts, as they tend to be pretty close, and possessions played.] of the proportion of a team’s offensive possessions end with a given player attempting to score either via a field goal attempt or a trip to the foul line. You might notice the positions don’t add up to 100%, that is because both of variation in lineups as well as the fact there are more turnovers than offensive rebounds, which means that the NBA as a whole averages just below one scoring attempt per possession. In any event:

PosScor
PosScor /

The smaller guys tend to shoot more, cue grumblings and not feeding the big fella enough and so on.

  • Playmaking

Playmaking Usage[2. Which I’ve sometimes referred to as “Assist Usage.”] is the biggest area neglected by traditional usage stats. Assist percentage gets us part of the way there, but the fact that Assist% is on a different scale as usage[3. And is indirectly affected by usage rates, as shots the player makes come out of the denominator in the equation for Assist%] makes meaningful comparison difficult. Plus, with SportVU data, we can look at assist chances rather than simply assisted shots made. Playmaking usage is simply assist chances[1. Including SportVU tracked “FT Assists” which I think are undercounted by the method feeding into the NBA.com data, but that’s another story for another time.] over possessions played.

Getting back to the study of positions, it’s not surprising point guards dominate on this front, with 2-guards picking up some of the slack.

PosPlay
PosPlay /

This large gap in playmaking shows the value of wing players who can close that gap. It’s not for nothing that I like to suggest players like LeBron James (20.2% Playmaking Usage), James Harden (18.7%) or even Gordon Hayward (14.4% in 2014/15, perhaps likely to rise in 2015/16 depending on how frequently Utah plays “no-PG” lineups with Dante Exum out for the year.) are filling the “Point Guard Role” on offense for their respective teams.

  • Turnovers

As stated above, traditional turnover rate is simply not a good stat. I’m not suggesting it’s been inaccurately calculated. Rather it does not provide the information it purports to. Specifically, it doesn’t do a great job of illuminating the players most and least prone to turnovers.  There are two things to look at with respect to turnovers: Turnover Usage[5. Much likes Scoring and Playmaking Usage, simply Turnovers / Possessions.] and what I’ve chosen to call “True Turnover Percentage.” This last is my suggested improvement on traditional TO% because it measures turnovers against both scoring and playmaking attempts, rather than simply scoring chances. Naturally, this gives the players who are at taking care of the ball[5. THAT’S HOW YOU GET CHOSEN TO PLAY POINT GUARD…at least in part.] a fairer shake than simply rewarding gunning and penalizing passing. By the numbers:

PosTO
PosTO /

So on a per possession basis, point guards do turn the ball over a little bit more. But controlling for their higher level of useful activity and it essentially equals out with other positions.[7. Aside from centers. Especially in today’s game, if they had good hands, they’d be power forwards instead, right?] Even controlling for playmaking possessions doesn’t quite do the ballhandling and protection skills of point guards their just due, which brings us to the last area, possession.

  • Possession

For positions no longer mattering they sure do seem to be relevant in terms of who has the ball:

PosToP
PosToP /

Admittedly, some of this possession is simply “who dribbles the ball up the floor.” However, against NBA level defense, that shouldn’t be taken for granted. It doesn’t happen often, but those times when a player, even a point guard, is simply unable to beat the ball pressure applied by his defender[8. A recent example might be Avery Bradley hounding the heck out of Ray Felton in the 2013 playoffs.] are uncomfortable in the same way as is the feeding scene in Jurassic Park.

The defense isn’t only trying to stop the offense from doing useful things, they are actively seeking to take the ball away. Protecting the ball isn’t just about not turning it over while trying to score or set up a teammate, and plenty of turnovers happen on simple ball-reversals, passes-fumbled out of bounds and moving screens not really involved in the play. Taking into account the need to keep a hold of the ball at all times is where the true wizardry of the NBA point guard can be seen. The following chart represents time of possession per turnover:

PosTTO
PosTTO /

It takes twice as long with the ball for a point guard to commit a turnover as it does a wing, and more than three times as long as for a big man. That’s why the ball stays in the hands lead guards as much as it does. Even if they aren’t actively doing much, they aren’t giving anything away either.

Putting it all together, here’s a graphical representation of the various usage categories and peripheral stats across positions:

That’s all for now. In the near future, I’m going to look at rebounding (a quick preview here), play types and anything else that might catch my eye.